SeaEval — leaderboard
Southeast Asian multilingual and cultural reasoning benchmark covering cross-lingual QA, cultural reasoning, safety, and regional language understanding.
Metric: Average Score (%). Source: huggingface.co. Status: saturation imminent. 30 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Qwen 2.5 72B Instruct | 72.61 |
| 2 | Qwen 2.5 32B Instruct | 70.56 |
| 3 | Qwen 2 72B Instruct | 70.53 |
| 4 | Qwen 2.5 14B Instruct | 69.6 |
| 5 | Llama 3.1 70B Instruct | 68.37 |
| 6 | Llama 3.1 70B | 67.78 |
| 7 | Llama 3 70B Instruct | 66.16 |
| 8 | Qwen 2.5 7B Instruct | 65.94 |
| 9 | Gemma 2 9B (IT) | 64.93 |
| 10 | Qwen 2 7B Instruct | 64.92 |
| 11 | SeaLLMs-v3-7B-Chat | 63.37 |
| 12 | Sailor2-8B-Chat | 59.95 |
| 13 | Qwen 2.5 3B Instruct | 59.13 |
| 14 | Llama 3.1 8B Instruct | 57.24 |
| 15 | Llama 3 8B Cpt Sea Lionv2.1 Instruct | 57.02 |
Interactive version: theaggregate.ai/benchmark?slug=seaeval · How the rankings work · Data refreshed daily, snapshot 2026-07-22.